There are many cases where the reaction forces are an amplification of the input forces. If you are thinking static equilibrium, then check all the reactions in X, Y, Z to confirm. My thought is that the X-sliding direction component of contact force is balanced to the input force, but that you looked at contact normal force instead. (magnitude, not X)
Keep in mind the flexing of your model is not really accounted for and so the model seems over-constrained because the bumps are trying to move the parts away from each other, yet they seem to be held by a slider.
It might be better to do FEA (with contact) rather than a mechanism so you can account for flex.
If your model has a low rate spring pushing the entire moving component against the stationary one, then mechanism could work. (local flex being negligible). For example if those two cylindrical holes hold springs to push the parts together then those need to be modeled to allow the part to move correctly. The slider would need to be removed and probably a planar constraint used. Creo also will have difficulty with contact from concave to convex surfaces, (most contact theory is convex-convex or convex-flat.)
Also, It seems the concave craters have sharp edges near where I think they will be contacting. Generally sharps should be avoided in contact areas so the progression is smooth. you could try a torroid to represent the fillets.
Anyway to make a long story short, it probably ain't as simple as it appears. Feel free to ask questions as you go
